Electrically Conductive Metal–Organic Frameworks for Electrocatalytic Applications

Metal–organic frameworks (MOFs), also known as porous coordination polymers (PCPs), first defined by Yaghi and coworkers in 1995, are a novel class of crystalline porous materials formed by the self-assembly of metal ions or clusters and organic ligands. MOFs are of considerable interest from many perspectives due to ultrahigh specific surface area, adsorption/absorption capabilities, postsynthetic modifications, high porosity, and tunable pore sizes and have been explored for a myriad of applications, including gas sorption, storage, drug delivery, luminescence, sensor technology, proton conduction, and heterogeneous catalysis. MOFs are composed of hard metal ions connected to redox-inactive organic ligands that are bound by hard oxygen or nitrogen atoms, causing the lack of low-energy pathways and free carriers for charge transport in MOFs. Therefore, most MOFs are inherently insulated (conductivity values lower than 10 10 S cm ), which largely limits their use in electrocatalytic systems. By imparting electrical conductivity to MOFs, their catalysis efficiency can be effectively improved. Conductive MOFs have a relatively short history, namely, from their preparation to utilization only a short decade. For a long time, the electrical properties of MOFs have received less attention, while their potential as electrically conductive porous materials has just been taken into account recently. According to different charge carriers, conductive MOFs can be classified into ionic conduction and electronic conduction. The earliest appearance of conductive MOFs can be traced back to 2009. Takaishi et al. reported the first electrically conductive MOFs, Cu[Cu(pdt)2] (pdt1⁄4 2,3-pyrazinedithiolate), which showed a high electrical conductivity of about 6 10 4 S cm 1 at 300 K. In the same year, Sadakiyo et al. synthesized a MOF with high proton conductivity, namely, (NH4)2(adp)[Zn2(ox)3]·3H2O (ox1⁄4 oxalate, adp1⁄4 adipic acid). This is the first MOF example exhibiting a superprotonic conductivity of 10 2 S cm 1 at ambient temperature. In 2010, Kobayashi et al. obtained Cu[Ni(pdt)2] using the samemethod as Cu[Cu(pdt)2]. The conductivity of Cu[Ni(pdt)2] is 1 10 8 S cm 1 and its Brunauer–Emmett–Teller (BET) surface area is 385m g , which is the first MOF that can take into account both high specific surface area and high conductivity. Subsequently, a variety of conductive MOFs were synthesized. Energy is indispensable in modern society. But traditional fossil energy sources such as coal, oil, and natural gas are nonrenewable and not environment friendly. So transforming our current energy system into a sustainable one is paramount. According to previous reports, both electrocatalysis and photocatalysis are effective for energy conversion and can satisfy the demands of green development. Thus, seeking electrocatalysts with high conversion rates and efficiency could be a useful way to L. Liu, Prof. Q.-L. Zhu State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter Chinese Academy of Sciences (CAS) Fuzhou 350002, China E-mail: qlzhu@fjirsm.ac.cn L. Liu, Prof. Q.-L. Zhu University of Chinese Academy of Sciences Beijing 100049, China Prof. Q. Xu Department of Materials Science and Engineering and SUSTech Academy for Advanced Interdisciplinary Studies Southern University of Science and Technology (SUSTech) Shenzhen 518055, China E-mail: xuq@sustech.edu.cn

mitigate environmental pollution and energy shortage. Conductive MOFs are considered as one of the ideal candidates for electrocatalysis due to their numerous active metal centers, large porosity, high conductivity, and structural rigidity which can facilitate surface contact and mass transfer. Recent years have witnessed a large number of high-performance conductive MOFs, and several even display activity comparable with that of the best heterogeneous catalysts. [21] Although many advances in the synthesis of electrically conductive MOFs have been made, their utilization in electrocatalysis still faces a series of challenges, such as the instability of most MOFs, the control of MOF film thickness, and the construction of electrodes. To achieve a real breakthrough for conductive MOF-based electrocatalysts, a focused review of the general approaches taken toward achieving high conductivities in MOFs and the electrocatalytic application of the conductive MOFs is necessarily imminent.
In this review, we discuss a set of electrically conductive MOFs, including intrinsically conductive MOFs, guest-based conductive MOFs, and conductive MOF composites, as well as their utilization in the field of electrocatalysis. The scope of this review is limited to electrically conductive MOFs, while the ionic conduction of MOFs, which have been reviewed elsewhere, [22][23][24] will be not discussed here. The purpose of this article is to provide a little guidance and inspiration to achieve the intriguing potential of electrically conductive MOFs in electrocatalysis. We hope this review can shed some light on the future development of this highly exciting area.

Conductive Mechanisms of MOFs
The conductivity (σ) of a conductive MOF is determined by the density (n) and mobility (μ) of electrons (e) and holes (h). [25] Therefore, to achieve high conductivity, it is essential to maximize both concentration and mobility of charge carriers (Equation (1)) σ ¼ eðμ e n e þ μ h n h Þ Most MOFs are intrinsically insulated due to the high barriers for charge transfer and the lack of free charge carriers. The materials with high concentrations of loosely bound charge carriers generally have high charge densities. The sources of charge carriers can originate from the metal ions, organic ligands, or guest molecules. And the pathways for charge transport can be briefly classified into two categories (Figure 1a): redox hopping transport and band transport. [26] In the former, charge carriers are localized at specific sites with discrete energy levels, and hop between neighboring sites. In the latter, the delocalized charge carriers can transport via continuous coordination or covalent bonds in MOFs. Both hopping transport and bond transport require a good spatial and energy overlap between orbitals of appropriate symmetry. Improving the charge transfer pathways by increasing orbital overlap enhances the charge mobility.
In general, there are two different design considerations in the synthesis of electronically conductive MOFs: 1) "through-bond" approach and 2) "through-space" approach ( Figure 1b). [26] Through-bond mechanism relies on improving covalent bonding between metals and ligands to achieve the high charge transfer in MOFs, while through-space mechanism harnesses noncovalent interactions between organic fragments.

Synthetic Strategies for Conductive MOFs
The field of conductive MOFs has witnessed a fast growth in the past 10 years. Many novel synthetic strategies have been developed to impart conductivity to MOFs at various levels. Using specific ligands and metal nodes to form long-range delocalized electrons for charge mobility is a rational choice to construct conductive MOFs. In addition, integrating guest molecules into MOFs or mixing MOFs with electrical conductors is also an intriguing strategy to tailor the scale of conductivity. According to different design strategies, conductive MOFs can be divided into three types: intrinsically conductive MOFs, guestbased conductive MOFs, and conductive MOF composites. Figure 1. Schematics of two charge-transport a) mechanisms and b) approaches, in MOFs. Reproduced with permission. [26] Copyright 2019, Royal Society of Chemistry.

Intrinsically Conductive MOFs
For intrinsically conductive MOFs, the conductivity is the inherent property of MOFs, rather than being induced by augmentations such as the exchange of guest molecules or conduction through the framework pores. Intrinsically conductive MOFs usually contain redox-active metal ions or ligands, or conjugated organic ligands that can generate charge transport pathways upon coordination with their metal ion counterparts. They are conductive due to their favorable spatial and energetic overlap between the metal and ligand orbitals as well as noncovalent interactions. A deliberate selection of metal ions and ligands can allow the synthesis of conductive MOFs.

Mixed-Valence Metal Ions and Ligands
The redox behavior of metal cations inside MOFs could provide a pathway for electrons. A vast number of conductive MOFs have been obtained by using metal ions with tunable valences as nodes. For instance, the first conductive MOF, which contains the donor Cu I and acceptor [Cu III (pdt) 2 ] À , is definitive evidence of the importance of the mixing valence. [15] A mixed-valence Cu(I)─Cu(II) framework was also reported by Okubo et al., which possesses low activation energy for charge transfer. [27] Similarly, the 3D framework [Cu I Cu II 2 (DCTP) 2 (NO 3 )]⋅1.5DMF (DCTP ¼ 4 0 -(3,5-dicarboxyphenyl)-4,2 0 :6 0 ,4 0 0 -terpyridine) also shows semiconducting behavior. [28] In addition, the conductive MOFs with mixed-valence iron ions have also been reported. Sun et al. systematically studied 20 MOFs with four different topologies, in which iron ion as the redox metal center endows the best electrical properties. [29] They proved that mixed-valence Fe II /Fe III increases the charge density and promotes charge delocalization, which leads to the high conductivity. In addition, the correlation between introduction of mixed-valence Fe II / III and increased conductivity has also been confirmed in Fe 2 (BDT) 3 (BDT ¼ benzene-1,4-ditetrazolate). [30] The conductivity of as-synthesized material was largely improved after 30 days in air, which is ascribed to the partial oxidation of Fe II centers to Fe III in the framework.
Mixed-valence ligands are also adopted to synthesize intrinsically conductive MOFs. Due to the partial oxidation of ligands in the synthesis, the resulting MOFs contain mixed-valent linkers, which may lead to high carrier density, thereby making MOFs conductive. Specific examples are a series of MOFs built with tetraoxolene ligands, such as dihydroxybenzoquinonate (dhbq) and chloranilate (Cl 2 dhbq). In 2015, Darago et al. synthesized [Fe 2 (dhbq) 3 ] 2À using dhbq as ligand. [31] The electrical conductivity of the as-synthesized material is up to 0.16 S cm À1 . The charge hopping between dhbq 3À• and dhbq 2À centers accounts for its such high conductivity. In 2017, the Harris group reported the conductivity of the Fe-chloranilate framework. [32] Due to the electron transfer of the Fe II ions to the ligands, the ligands in the framework are in mixed valency (Cl 2 dhbq 3À• , Cl 2 dhbq 2À ). It shows a high conductivity of 1.4(7) Â 10 À2 S cm À1 . When all ligands are converted to the Cl 2 dhbq 3À• radical state, the conductivity decreases to 5.1(3) Â 10 À4 S cm À1 . The relatively low conductivity (1.5(3) Â 10 À9 S cm À1 ) of [Zn 2 (Cl 2 dhbq) 3 ] 2À also illustrates that mixed valency plays an important role in the conductivity.

Infinite 1D/2D Metal-Sulfur Chains
The conductivity of MOFs could be realized by forming infinite 1D/ 2D metal-sulfur chains. Sun et al. reported a sulfur-substituted Mn-MOF-74 analogue Mn 2 (DSBDC) (DSBDC ¼ 2,5-disulfhydrylbenzene-1,4-dicarboxylic acid) with infinite 1D Mn─S zigzag chains along the crystal c axis, as shown in Figure 2a. [33] This material remains permanently porous, with a BET surface area of 978 m 2 g À1 . They subsequently reported its electrical conductivity of 2.5 Â 10 À12 S cm À1 , which is about one order of magnitude higher than that of Mn 2 (DOBDC) (DOBDC ¼ 2,5-dihydroxybenzene-1,4-dicarboxylic acid) (3.9 Â 10 À13 S cm À1 ). [34] The result indicates that (─Mn─S─) ∞ chain is more effective than (─Mn─O─) ∞ chain in charge transfer, which may be attributed to the better energy matching between the frontier orbitals of the thiolate and the metal, thus increasing the conductivity of Mn 2 (DSBDC). They also studied the conductive difference of Fe and Mn analogues and found that the conductivity of Fe 2 (DSBDC) (3.9 Â 10 À6 S cm À1 ) Figure 2. a) Two conductive MOFs with infinite (─M─O─) ∞ and (─M─S─) ∞ chains. Reproduced with permission. [33] Copyright 2013, American Chemical Society. b) A conductive MOF with infinite (─Cu─S─) n chains. Reproduced with permission. [35] Copyright 2019, Nature Publishing Group. is about 1 million-fold higher than that of Mn 2 (DSBDC), and also much higher than that of Fe 2 (DOBDC) (3.2 Â 10 À7 S cm À1 shows a high conductivity up to %11 S cm À1 , which is originated from the integration of the 2D (─Cu─S─) n plane in the structure (Figure 2b). [35] Infinite metal-sulfur chain is again invoked to explain the higher conductivity.

π-Interactions
According to the "through-space" approach, the efficient charge transport can also occur in MOFs through noncovalent interactions, particularly π-π stacking. Frameworks based on the well-known electron donor tetrathiafulvalene (TTF) constitute a large portion of MOFs designed following this principle. In 2012, Narayan et al. synthesized Zn 2 (TTFTB) using tetrathiafulvalene-tetrabenzoate (H 4 TTFTB) and Zn 2þ , whose charge mobility is commensurate with some best organic semiconductors. [36] As shown in Figure 3a,b, the material possesses both columnar stacks of TTF and permanent pores lined by benzoate linkers, and a relatively close intermolecular S···S contact of 3.803(2) Å is found between neighboring TTF moieties, which is similar to the range of intermolecular S···S distances found in TTF-TCNQ (TCNQ ¼ 7,7,8,8-tetracyanoquinodimethane) and other conductive charge-transfer salts. Park et al. in 2015 fabricated a series of isostructural M 2 (TTFTB) with M ¼ Mn, Co, Zn, and Cd. [37] All four MOFs contain infinite 1D π-π stacked helical TTF and their electrical conductivities range from %10 À6 to 10 À4 S cm À1 , which coincides with the size change of the metal cations. Cd 2 (TTFTB) has the largest cation size and shows the highest conductivity of 2.86 Â 10 À4 S cm À1 , while the conductivities of Zn 2 (TFTB) and Co 2 (TTFTB) are only 3.95 Â 10 À6 and 1.49 Â 10 À5 S cm À1 , respectively. The authors proved that their electrical conductivities are determined by the S···S distances between neighboring TTF moieties ( Figure 3c). As the cation radius increases, the S···S distance decreases from 3.77 Å for Co 2 (TTFTB) to 3.65 Å for Cd 2 (TTFTB), resulting in better orbital overlap and higher conductivity. In 2018, Xie and Dincȃ reported several electrically conductive MOFs based on H 4 TTFTB and lanthanide ions. [38] They confirmed that the S···S contacts between ligands provide possible charge transport pathways in the materials. Later, they synthesized three TTFTB-MOFs based on tetrathiafulvalene linker . Reproduced with permission. [36] Copyright 2012, American Chemical Society. c) Correlation between S···S distance and electrical conductivity in M 2 (TTFTB) (M ¼ Mn, Co, Zn, and Cd). Reproduced with permission. [37] Copyright 2015, American Chemical Society. d) Correlation between S···S distance and electrical conductivity (batch and average values) in La 4 (HTTFTB) 4 (1), La(HTTFTB) (2), and La 4 (TTFTB) 3 (3). Reproduced with permission. [39] Copyright 2019, Royal Society of Chemistry.

Guest-Based Conductive MOFs
The structures of MOFs derived from connected metal ions and linkers endow them with a large number of available pores, which present both a challenge and an opportunity with respect to conductivity. The incorporation of guest molecules into the inherent porosity is one of the most frequently adopted strategies to improve the conductivity of MOFs. These guest molecules, which can be introduced during the preparation of MOFs or by postsynthetic modification (PSM), can interconnect with the secondary building units (SBUs) of MOFs, thus generating redox active sites inside the network. Redox-active molecules, metal oxides, metal nanoclusters, cationic species, and conducting polymers are the widely used dopants in recent years.
www.advancedsciencenews.com www.advenergysustres.com Later, the metallocarborane molecule, nickel (IV) bis(dicarbollide) (NiCB), was incorporated into MOF channels as the guest. A study by Kung et al. displayed that NiCB was selectively loaded in the micropores of NU-1000 while leaving the mesopores unoccupied, as shown in Figure 7b. [73] By modifying, the conductivity of NU-1000 MOF increases from undetectable to 2.7 Â 10 À7 S cm À1 . UV-vis spectroscopy data revealed charge transfer between the pyrene-based linker and NiCB, which endows the material with electrical conductivity. In addition, the effects of tetratin(IV)oxy clusters doping on the conductivity of NU-1000 have also been investigated. [74] By installing tetratin(IV)oxy clusters, NU-1000 obtains a significant improvement in conductivity, with the value as high as 1.8 Â 10 À7 S cm À1 .
Guo et al. demonstrated that cationic methyl viologen (MV 2þ ) species can also be used to prepare conductive MOFs (Figure 7c). [75] They soaked films of a pillared paddlewheel MOF with an electron-rich pyridyl-and dipyrrolidyl-substituted naphthalene diimide (NDI) ligand in a solution of MV 2þ , resulting in an enhanced conductivity, from 6 Â 10 À7 to 2.3 Â 10 À5 S cm À1 .

Conducting Polymers
The guest-promoted approach is not limited to small species, and large conducting polymers have also been reported to increase the conductivity by polymerization in the pores of MOFs. [76,77] For MOFs⊃polymer assemblies, the conductivity of the resulted MOFs is higher than either of the pristine components. Enhanced transport properties in MOFs⊃polymer materials have been attributed to charge transfer interactions between polymer chains surrounded by π-donor ligands, as well as high degrees of order and orientation in the polymers. [78] Several common conducting polymers such as polyaniline, poly-3,4-ethylenedioxythiophene (PEDOT), polythiophene, and polypyrrole have been intercalated into different MOFs to improve their conductivities. For instance, Le Ouay et al. obtained a conductive porous composite material composed of Cr-MIL-101 and PEDOT, which was prepared by the polymerization of EDOT monomer in the cavities of MOFs with the catalysis of I 2 (Figure 7d). [79] The conductivity of MIL-101(Cr)⊃PEDOT is 1.1 Â 10 À3 S cm À1 , much higher than that of native Cr-MIL-101(10 À11 S cm À1 ). Recently, Mohmeyer et al. reported that the composite of Zr-bzpdc-MOF and PEDOT also exhibits good conductivity. [80] Similarly, NU-1000 is given conductivity through the incorporation of a designed pentathiophene derivative by solventassisted ligand incorporation, as shown in Figure 7e. [81] By varying the doping level of the polythiophene, the material could be changed from an insulator to a semiconductor. These findings provide an approach for fabricating conductive porous materials and expand opportunities for applications of MOFs. Other instances of polypyrrole and polyaniline incorporation resulting in conductivity have also been reported. [82][83][84][85][86][87][88][89]

Conductive MOF Composites
Hybridization of MOFs with carbon-based materials (like carbon nanotubes (CNTs), graphene oxide, graphene, quantum dots) is a valuable strategy to render or facilitate conductivity in the resulting composites. Furthermore, the growth of MOF nanostructures (nanosheets, nanorods, nanowires, etc.) onto conductive substrates is another way to allow the extension of the electrochemical applications of MOFs.

MOF-Carbon Composites
Goswami et al. reported that the conductivity of C 60 @NU-901 ( Figure 8a) has been greatly improved by physically encapsulating an excellent electron acceptor, C 60 , into the channels of NU-901, from unmeasurable to 10 À3 S cm À1 . [90] The diameter Figure 8. a) Immobilization of C 60 within diamond shaped channels of NU-901. Reproduced with permission. [90] Copyright 2018, Royal Society of Chemistry. b) Synthesis of GO/ZIF-8 and RGO/ZIF-8 nanocomposites with hierarchical porosity. Reproduced with permission. [92] Copyright 2016, Royal Society of Chemistry.
www.advancedsciencenews.com www.advenergysustres.com of C 60 (%7 Å) is well suited for encapsulation in NU-901 with pore aperture of 12 Å. Such remarkable conductivity enhancement is attributed to efficient charge-transfer interactions with pyrene moieties and C 60 acting as π-donor and acceptor. Graphene and graphene oxide have also been developed to construct conductive MOFs composites due to their wonderful electron mobility. For example, a hybrid material of graphene-like oxidized carbon nanoparticles and insulating HKUST-1 exhibits a tunable electrical conductivity. [91] Preliminary electrical measurements showed that the direct-current conductivity in the samples improves greatly with the increase in carbonaceous layer content. The authors estimated that the order of the conductivity could increase to 10 À4 S cm À1 when the carbon percentage is 40%. Furthermore, the sample of RGO/ZIF-8 (RGO ¼ reduced graphene oxide, 20 wt%) displays a conductivity of 0.64 S cm À1 , which is the highest among all 3D microporous MOFs, while the GO/ZIF-8 is nonconductive due to the insulating properties of both GO and ZIF-8 (Figure 8b). [92]

MOF-Conductive Substrate Composites
MOF-conductive substrate composites are in situ synthesized on conductive substrates without binder materials, and they have been widely used in electrocatalytic reactions due to the ability of conductive substrates to enhance the conductivity of loaded MOF samples. Although no conductivity data are available in most reports, MOFconductive substrate composites usually exhibit extraordinary activity for electrochemical reactions. One of the most commonly conductive substrates is nickel foam (NF). [93][94][95][96][97] For example, Sun et al. reported a binder-free 3D electrode, MIL-53(FeNi)/NF, which shows high current density (50 Ma cm À2 ) at an overpotential of 233 mV, a Tafel slope of 31.3 mV decade À1 , and remarkable stability for oxygen evolution reaction (OER). [94] Similarly, the Co-MOF nanoarrays grown on NF displays extraordinary OER catalytic activity and superior stability. [95] An interfacial engineering approach which allows preferred chelation of carboxyl groups in the ligands with the metal interlayers was used to synthesize the electrocatalytic materials. Nickel-iron foam (NFF) is also an ideal substrate material due to its high air permeability and high specific surface area. Cao et al. prepared a self-supporting MOF-based nanocomposite electrode (NiFe─NFF) by using the NiFe alloy foam as the semisacrificial template. [98] As shown in Figure 9, NFF was used as both metal sources and substrate. The results show that the bimetallic MOF composite NiFe─NFF possesses remarkable electrocatalytic OER activity with required overpotentials of 227 and 253 mV to achieve current densities of 10 and 100 mA cm À2 , respectively. Lin et al. synthesized Ni-MOF/Ni/CC by in situ growth of ultrafine and ultralong Ni-MOF nanowire arrays on highly rough and conductive scaffolds of porous Ni/CC. [99] The resultant composite has an amazing water oxidation activity and durability with a η 10 of 240 mV at the end of a constant current electrolysis for 32 h. In addition, according to previous reports, fluorine tin oxide (FTO) has also been used to construct conducting MOFs composites. [100,101]

Conductive MOFs for Electrocatalysis
One of the most appealing applications of conductive MOFs is as electrocatalysts due to their uniform active sites, high porosity, and enhanced conductivity. In addition to the intrinsic electrical conductivity, the morphology of MOFs plays a significantly important role in electrocatalysis. Particularly, constructing the thin 2D nanostructures of MOFs could greatly expose the active sites and further enhance electrical conductivity due to the nanosized thickness, opening up an opportunity to break through the bottleneck of MOF-based electrocatalysis. Moreover, the stability of conductive MOFs in electrocatalysis (strong acidic/alkaline solutions) is also a crucial factor. Although many conductive MOFs have been synthesized, the research on their electrocatalytic applications is obviously lagging behind. In this section, we briefly summarize some successful examples of conductive MOFs used as electrocatalysts for the most important electrochemical reactions, including hydrogen evolution reaction  (HER), OER, oxygen reduction reaction (ORR), and CO 2 reduction reaction (CO 2 RR).

HER
Metal-dithiolene-based MOFs have been extensively studied as promising electrocatalysts for HER. For example, Ni-THT, Co-BHT, and Co-THT have been reported to display remarkable electrocatalytic activities for HER. [57,58] The thin-film Ni-THT samples were prepared through an air-liquid interface process (the Langmuir-Blodgett method). The Tafel slope of Ni-THT is 80.5 mV decade À1 and the overpotential is 333 mV at 10 mA cm À2 . Both Co-BHT and Co-THT exhibit impressive stability under acidic conditions and the Tafel slopes are between 149 and 189 mV decade À1 at pH ¼ 4.2, and exchange current densities of 10 À5.3 A cm À2 . The overpotentials are 340 and 530 mV, respectively, at pH ¼ 1.3 and a current density of 10 mA cm À2 . In 2017, a report showed that Cu-BHT nanoparticles have good activity and stability for HER with a Tafel slope of %95 mV decade À1 and an overpotential of 450 mV at 10 mA cm À2 under optimized conditions. [102] Downes and Marinescu reported a series of metal-dithiolene-based MOFs using benzene-1,2,4,5-tetrathiolate (BTT) and benzene-1,2,4,5-tetraselenolate (BTSe) as ligands and all of them can be used as good hydrogen-evolution electrocatalysts. [103][104][105] In 2018, the same group reported the evaluation of the HER activity of Co-BHT, Ni-BHT, and Fe-BHT. [106] The reaction was conducted in an acidic aqueous solution (pH ¼ 1.3) and the overpotentials of BHT-based MOFs follow the order of Fe-BHT (405 mV) > Ni-BHT (331 mV) > Co-BHT (185 mV). The lower overpotential of optimized Co-BHT should benefit from its improved charge transfer properties. They also studied the effect of film thickness on H 2 production and found that as the film thickness was increased, the overpotential initially decreases because thicker films correspond to higher bulk catalyst loadings for layered materials, so more accessible active sites can be realized. But when increased to a certain thickness, electrons and protons had poor diffusion in the thick film, which inhibited HER activity. In addition, metal-dithiolene-diamine MOFs have also been investigated as electrocatalysts toward HER. [107] Polyoxometalate-based MOFs (POMOFs) also show outstanding electrocatalytic efficiency for HER. In 2011, Nohra et al. prepared three novel POMOFs under hydrothermal conditions and used them for HER. [108] They found that POMOF-based electrodes exhibit higher activity than platinum with a hydrogen yield of more than 95% and a turnover number of 1.2 Â 10 5 after 5 h. In 2015, another POMOF (NENU-500) with good tolerance to acidic and alkaline media was found to play superior HER performance. [109] It shows an onset overpotential of 180 mV and a Tafel slope of 96 mV decade À1 . Moreover, the electrocatalytic activity of NENU-500 was maintained after 2000 cycles.

ORR
The first report on the use of MOFs as electrocatalysts for ORR is Cu-BTC (Cu 3 (BTC) 2 ). [126] However, Cu-BTC is structurally unstable in aqueous media. A water-stable MOF, copper www.advancedsciencenews.com www.advenergysustres.com (II)-2,2 0 -bipyridinebenzene-1,3,5-tricarboxylate (Cu-bipy-BTC, bipy ¼ 2,2 0 -bipyridine) was thus prepared and can catalyze the ORR with almost 4e À transfer pathway. Subsequently, another Cu-MOF was reported to be used as ORR electrocatalyst. [127] The MOF layer on RGO immobilized glassy carbon electrode can catalyze the ORR through a 2─4 electrons reduction pathway. As an intrinsically conductive MOF, Ni 3 (HITP) 2 has been demonstrated as a well-defined, tunable oxygen reduction electrocatalyst. [128,129] Its catalytic activity for ORR is competitive with the most active nonplatinum group metal electrocatalysts. Transition metal porphyrins are found to be highly efficient molecular catalysts for ORR. Accordingly, PCN-223-Fe, which is constructed from Zr 6 -oxo clusters and Fe(III) porphyrin linkers, has been observed to catalyze ORR through the nearly 4e À pathway with high selectivity. [130] Composite materials composed of MOFs and conductive supports are demonstrated to facilitate the ORR. In 2012, a graphene-metalloporphyrin MOF composite was fabricated and used for ORR. [131] It was found that the addition of pyridine-functionalized graphene enhanced the electrochemical charge transfer rate of iron-porphyrin; thus, the hybrid MOF shows an improved catalytic activity and facile 4-electron ORR. Co-MOF@CNTs was reported to be a superior bifunctional electrocatalyst due to the synergistic catalysis of Co(II), organic ligands, and CNTs. [132] It shows outstanding stability and notable OER and ORR catalytic activities, which are comparable with RuO 2 and 20 wt% Pt/C catalysts, respectively.

CO 2 RR
Reducing carbon dioxide to valuable chemicals has always been a hot issue. [133,134] The most commonly used catalysts for CO 2 RR are Cu-based MOFs. [135][136][137] For instance, the uniform Cu 3 (BTC) 2 film, supported on glassy carbon electrode, was tested for CO 2 RR in CO 2 saturated DMF solution in the presence of supporting electrolyte. [138] The major product oxalic acid (H 2 C 2 O 4 ) was obtained with 90% purity and the Faradaic efficiency (FE) of 51%. Hinogami et al. also demonstrated the high selectivity of Cu-based MOFs. [139] They prepared a copper rubeanate MOF (CR-MOF) to reduce CO 2 into HCOOH with the selectivity of more than 98%, while the Cu metal electrode generated a series of products, including formic acid, methane, ethylene, and ethane. In another report, the authors investigated four Cu-based metal-organic porous materials for CO 2 RR, which have high efficiency for the production of methanol and ethanol in the liquid phase. [140] It is due to their large surface areas, accessibilities of the electrodes as well as exposure of the Cu catalytic centers.
www.advancedsciencenews.com www.advenergysustres.com of CO 2 to CO in an aqueous electrolyte. [142] The selectivity for CO production exceeds 76% and the stability is more than 7 h with a turnover rate (TON) of 1400 per site. In situ spectroelectrochemical measurements showed that the majority of catalytic centers in this MOF are redox-accessible where Co(II) was reduced to Co(I) during catalysis. So it might be an effective method for electrochemical CO 2 reduction with excellent performance to integrate active species into MOFs. This was also confirmed by Wang et al., who reported the ZIF-8 doped with electrocatalytically active 1,10-phenanthroline. [143] The experimental results indicate that the electron-donating nature of phenanthroline promotes electron transfer, which facilitates the formation of *COOH, thus improving the activity and FE toward CO production.

Conclusions and Perspectives
Owing to myriad intriguing and desirable physicochemical properties, MOFs are potential for a wide range of applications, especially in catalysis. However, the lack of electrical conductivity largely restricts their adaptation as electrocatalysts. The emergence of conductive MOFs is a good solution to this problem. In recent years, many effective approaches have been successfully developed to engender conductivity of MOFs, which is beneficial to broaden their electrical applications. Mixed-valence metal ions or ligands, infinite 1D/2D metal-sulfur chains, and π-interactions can be used to synthesize intrinsically conductive MOFs. The introduction of redox-active molecules, metallic species as well as conducting polymers can construct guest-based conductive MOFs. Furthermore, MOF-carbon composites and MOF-conductive substrate composites also show considerable conductivity. Although these methods greatly enrich the library of conductive MOF, their application in electrocatalysis is still quite limited. The vulnerability of most MOF materials is a dominant factor. Therefore, improving the stability of electrocatalysts in strongly acidic and alkaline solutions to avoid hydrolysis and framework collapse is of vital importance to put conductive MOFs into practice. In addition, the thickness of MOF film also greatly influences the catalytic activity. The use of adhesives during electrode construction also hinders charge transfer and thus reduces electrocatalytic activity. So exploiting effective methods to optimize MOF deposition on the electrode or substrate is essential for their utilization in electrochemical fields.
As the shining stars in the field of materials science, though faced with a number of key challenges, a few successful examples of conductive MOFs with fantastic catalytic efficiency reveal their promising application of electrocatalysis. Given the great efforts and encouraging advances in the last few years, the field of conductive MOFs is growing, and more and more conductive MOFs are used in electrocatalysis, some even outperform the most active catalysts. We are confident that further development will constantly provide opportunity for conductive MOFs to be served as novel electrocatalysts.